M.Sc. Pascal Maierhofer

B.Sc. Onur Ünal

Motivation

Modeling and simulating patient-specific heart geometries hold great potential for helping electrophysiologists identify the origins of atrial fibrillation (AF) and accelerate clinical interventions. However, creating these models from clinical data remains a challenging and labor-intensive process due to anatomical variations among patients. To facilitate the creation of personalized models, AugmentA [1] was developed at IBT. AugmentA automates the annotation based on specific rules for each atrial region and helps with further data augmentation. However, it still encounters difficulties with certain patient geometries, further limiting the already sparse amount of available clinical data.

Task

This project will focus on refactoring, documenting, and enhancing the existing codebase and workflow of AugmentA. The goal is to identify weaknesses in the current code, improve its readability and stability, and optimize the algorithms currently in use. You can have a look at the current version at https://github.com/KIT-IBT/AugmentA.